pytho脚本通过boto3访问radosgw提供的对象存储

经常我们需要测试Ceph的RadosGW是否访问正常，s3cmd是一种方法，另一个就是用各种s3的SDK了，这里介绍python的boto3 SDK；

脚本如下：

$ cat bucketops.py

#!/usr/bin/env python

# coding=utf-8

import boto3

def main():

    access_key = 'IYZ800MD*******'

    secret_key = '05HfpRrrh1Gs1p8bx********'

    s3_host = 'http://'

    bucket_name = 'bruins'

    object_key = 'hello.txt'

    s3client = boto3.client('s3',

                        aws_secret_access_key = secret_key,

                        aws_access_key_id = access_key,

                        endpoint_url = s3_host)

   response = s3client.list_buckets()

    for bucket in response['Buckets']:

            print "Listing owned buckets returns => {0} was created on {1}\n".format(bucket['Name'], bucket['CreationDate'])

    # creating a bucket

    response = s3client.create_bucket(Bucket = bucket_name)

    print "Creating bucket {0} returns => {1}\n".format(bucket_name, response)

    # creating an object

    response = s3client.put_object(Bucket = bucket_name, Key = object_key, Body = 'Hello World!')

    print "Creating object {0} returns => {1}\n".format(object_key, response)

    hello_url = s3client.generate_presigned_url('get_object', Params={'Bucket': bucket_name, 'Key': object_key}, ExpiresIn= 3600)

    print hello_url

if __name__ == '__main__':

    main() 

注意的问题：

最后一步生成的object url，通过wget访问时需要把一些特殊字符进行转义；